Menu Close

//if (session_status() !== PHP_SESSION_ACTIVE) {session_start();}
//echo session_id();

Success! After launching the simulation on the previous page, IMASCOD performed the following operations:

  • the simulation of the source and background noise light rays propagation in the ECLAIRs coded mask telescope, resulting in an image recorded by the detector;
  • from this recorded image, the reconstruction of the sky image by the deconvolution process (explained below). You will see your source appear, if you have given it enough intensity to stand out from the background noise.


Principle of the light rays propagation in the ECLAIRs telescope

Figure 1 illustrates the principle of coded mask imaging used by ECLAIRs. Light rays from a point source in the sky arrive parallel to the telescope (orange) and project a portion of the coded mask on the detector. For each position of the source chosen in the sky, the portion of the mask projected on the detector is different. There are two situations:

  • when the projection of the mask covers the entire detector: the source is said fully coded (left in Figure 1);
  • when the projection of the mask covers only a part of the detector (while the other part of the detector is behind the shielding seen from the source): the source is said partially coded (right in Figure 1).

Figure 1: depending on the position of the source in the sky, the whole detector is impacted (total coding, left) or only part (partial coding, right). The photons of the isotropic background noise in the sky impact the entire detector. For the same photon flux, a fully encoded source will be better detected than a partially coded source.

Projection of your source through the coded mask

For your chosen source position, IMASCOD generated Figure 2 below, which illustrates the projection of the source through the coded mask.

  • The portion of the mask projected on the detector for the position of the chosen source is indicated in yellow (on the left in Figure 2). If the source is fully coded, the yellow portion is the size of the detector. The closer the source is to the edge of the field of view, the weaker its coding will be, and the smaller the yellow portion projected on the detector will be (try for example x = 70, y = 130).
  • The image recorded by the detector (on the right in Figure 2) shows the pattern corresponding to the projected portion (in yellow) of the coded mask. For each position of the source, we obtain a different pattern; it is this pattern that uniquely “codes” the direction of the source.

Figure 2: Left: Pattern of the coded mask (the part opaque to photons is in black, the transparent part is in white) and its portion projected on the detector (in yellow). Right: pattern recorded by the detector (the color represents the number of photons of the source received by each pixel).

Addition of the background noise

In addition to the photons of the source, other photons of the diffuse background noise of the sky pass through the telescope mask. The latter come from all directions isotropically and illuminate the entire detector.
The image recorded by the detector (Figure 3) includes both photons of the source and background noise. Depending on the relative intensity of the source and noise, in some cases it is no longer easy (or impossible) to recognize the pattern of the mask projected by the source. The sky image reconstruction process by deconvolution (see following paragraph) nevertheless makes it possible to highlight the source in many cases.

Figure 3: image recorded by the detector. The color represents for each pixel of the detector the number of photons (of the source and the noise) detected during a given exposure time.

Sky image reconstruction

The sky image is reconstructed by IMASCOD from the image of the detector plane by a mathematical operation called deconvolution. The result is shown in Figure 4. If you have projected enough photons from your source on the detector, you will find it at the position you chose.

The principle of the reconstruction algorithm using deconvolution, illustrated in the video below, is to “digitally” drag the array of the coded mask “above” the table representing the detector image (Figure 3) to obtain one to one the pixels of the sky image (Figure 4). The source is found at the sky position where the correlation is the best.

On the video, a simple calculation is made between the mask and detector pixels in their intersection region. The result of this operation gives the correlation (the resemblance) between the mask pattern and the projected pattern. The larger the number is, the more similar the two are. The source is found where the resemblance is maximal. For deconvolution on board the satellite, the calculation involving the pixels of the mask and the detector in the region of the intersection is a little more complicated and makes it possible to improve the localization and computing time performances.
In the case of several sources, whose projections are mixed on the detector image, the algorithm is able to bring them out. The off-line version of IMASCOD is used to illustrate the mathematical formulas for deconvolution, and also deals with the case of several point sources.

Figure 4: sky image reconstructed by deconvolution. For each pixel in the sky, the color represents the reconstructed signal-to-noise ratio (in number of standard deviations or “sigmas”).

Onboard the SVOM satellite, the trigger software for detection of gamma-ray bursts embedded in the Scientific Processing Unit of ECLAIRs, reconstructs in a similar way to IMASCOD the sky images from the images recorded by the detector. This process is repeated in a methodical way in order to detect, at any time, the appearance of a new transient source (lasting between 10 ms and 20 min), at any position in the field of view and so to launch the alert as quickly as possible: a gamma burst has been detected!

The ECLAIRs trigger is at the heart of the SVOM mission thanks to the gamma-ray burst alerts it generates in real time. They drive to the automatic slew of the satellite to observe the afterglow of the burst with the onboard MXT and VT telescopes. In parallel they are relayed to GWACs and GFTs on the ground, as well as to all interested scientific observers.

More information on the ground-space strategy…


Related Posts